Privacy-preserving friendship establishment based on blind signature and bloom filter in mobile social networks

Abstract

Recent years have witnessed the rapid advancement of Mobile Social Networks (MSNs), in which mobile users can make friends and conduct other social activities. Common Friends is a popular application that allows a mobile user to participate in a matching process for mutual friends discovery, utilizing the information of participates' friends. However, privacy of participants and their friends may be breached if the matching protocol is not carefully designed. Although several protocols based on Private Set intersection (PSI) have been proposed to protect users' privacy, the complicated underlying cryptographic tools or interactive processes make them infeasible to be implemented in resource-restricted mobile devices. In this paper, based on blind signature and bloom filter, we propose an efficient privacy-preserving friendship establishment protocol, which can ensure users' privacy and reduce the complexity of set intersection computation. We introduce a bloom filter on the responder's side to improve efficiency of mobile users' friends information comparison and employ blind signature to prevent the initiator's input from being inferred by malicious responders with large amount of enumeration. The security analysis and performance evaluation demonstrate that our protocol works well for users' privacy protection and is feasible to be implemented.